Sharif Digital Repository

A coupled statistical and conditional generative adversarial neural network is used for 3D reconstruction of both homogeneous and heterogeneous porous media from a single two-dimensional image. A statistical approach feeds the deep network with conditional data, and then the reconstruction is trained on a deep generative network. The conditional nature of the generative model helps in network stability and convergence which has been optimized through a gradient-descent-based optimization method. Moreover, this coupled approach allows the reconstruction of heterogeneous samples, a critical and serious challenge in conventional reconstruction methods. The main contribution of this work is to... 

In this work, a novel approach to model the dynamic behavior of the flash separation process (as a main building block of non-reacting stage-wise operations) based on Mamdani Fuzzy Inference Systems is proposed. This model surmounts the need to solve various types of mathematical equations governing the system and does not require thermodynamic properties which are either not available or computationally demanding. Hence it can be easily used in dynamic simulation of multi-phase flow in distributed systems. In the proposed approach the overall model is broken into several simple sub-models based on intuitive analysis of an expert. Moreover, a new fuzzy concept, named “Linguistic Composition... 

A new two-dimensional dynamic model was developed to simulate asphaltene precipitation, aggregation, and deposition at isothermal and non-isothermal conditions. The perturbed-chain statistical associating fluid theory equation of state was used to model the asphaltene precipitation. Also, novel kinetic models were used to account for the aggregation and deposition of asphaltene particles. The effect of the aggregate size on the rate of aggregation and deposition was studied, and it was concluded that the rate of asphaltene deposition increases, while the concentration of nanoaggregates increases in the well column. The tendency of smaller aggregates to deposit on the surface could be... 

In this paper we propose a method towards purity control of pressure swing adsorption (PSA) processes which is based on the use of hybrid systems formalism. Hybrid systems feature both continuous and discrete-event dynamics and hence are very suited to describe in detail PSA processes. Based on mechanistic model of the processes, a local reduced-order model (LROM) is developed for PSA processes. Then the processes are represented as hybrid systems whose continuous evolution is described by the LROM. We then perform an analysis of hybrid reachability properties of the hybrid system obtained, based on which the so-called maximal safe set is computed. The analysis is performed for a two-bed,... 

Object-oriented modeling methodology is used for encapsulating different methods and attributes of data reconciliation (DR) in classes. Classes which are defined for DR, cover steady-state, dynamic, linear and nonlinear DR problems. Two main classes are Constraints and DR and defined for manipulating constraints and general DR problem. The remaining classes are derived from these two classes. A class namely DDRMethod is developed for encapsulating all common attributes and methods needed for any DDR method. Developed DR software and the method of performing dynamic DR are discussed in this paper. Two illustrative examples of Extended Kalman Filtering and artificial neural networks are used... 

This paper is aimed at investigating some computational aspects of different isoperimetric problems on weighted trees. In this regard, we consider different connectivity parameters called minimum normalized cuts/isoperimetric numbers defined through taking the minimum of the maximum or the mean of the normalized outgoing flows from a set of subdomains of vertices, where these subdomains constitute a partition/subpartition. We show that the decision problem for the case of taking k-partitions and the maximum (called the max normalized cut problem NCP^M), and the other two decision problems for the mean version (referred to as IPP^m and NCP^m) are NP-complete problems for weighted trees. On... 

This paper concerns a study on the optimal control for nonlinear systems. An appropriate alternative in order to alleviate the nonlinearity of a system is the exact linearization approach. In this fashion, the nonlinear system has been linearized using input-output feedback linearization (IOFL). Then, by utilizing the well developed optimal control theory of linear systems, the compensated nonlinear system has been controlled. Thus, the structure of the objective function will be converted into a quadratic form which is qualitativly comparable with usual cost functions, and from operating viewpoint is more favorable. To qualify such a procedure, it has been applied to two minimum and... 

Fuzzy sliding mode control (FSMC) as a robust and intelligent nonlinear control technique is proposed to control processes with severe nonlinearity and unknown models. The performance of the proposed method has been evaluated for both single input single output (SISO) and MIMO nonlinear systems through its application in three severely nonlinear processes that are frequently used as benchmarks of nonlinear process control strategies. The evaluation shows that, despite its lack of dependence on the process model, the proposed method performs almost the same as conventional sliding mode control alternatives that utilize all the information that exists in the mathematical model of the process.... 

An adaptive input-output linearization method for general nonlinear systems is developed without using states of the system. Another key feature of this structure is the fact that, it does not need model of the system. In this scheme, neurolinearizer has few weights, so it is practical in adaptive situations. Online training of neurolinearizer is compared to model predictive recurrent training. Relationships between this controller and neural network based model reference adaptive controller are established. A CSTR reactor and pH control in a neutralization process illustrate performance of this method. Simulation studies show a superior performance with respect to a PI controller  

Kinetic modeling is an important issue, whose objective is the accurate determination of the rates of various reactions taking place in a reacting system. This issue is a pivotal element for the process design and development particularly for novel processes which are based on reactions taking place between various types of species. In this paper, the Genetic Algorithms have been used to develop a systematic computational framework for kinetic modeling of various reacting systems. This framework can be used to find the optimum values of various parameters that exist in the kinetic model of a reacting system. The Fischer-Tropsch (FT) reactions have been used as the kinetic modeling bench... 

We proposed a personalized bolus advisor for patients with type 1 diabetes (T1D). A bolus advisor is a decision support system that recommends insulin doses based on an open-loop model-based optimization. To construct the bolus advisor, the optimal open-loop control of blood glucose (BG) concentration in T1D patients was represented as a multi-objective optimization problem. The insulin types, doses, and times for each injection were provided by the bolus advisor based on a personalized model and an average daily diet, which should be re-tuned frequently in specific time intervals. The constructed personalized model for T1D patients incorporates effects of the patient's age and body weight.... 

The bronchial tree plays a main role in the human respiratory system because the air distribution throughout the lungs and gas exchange with blood occur in the airways whose dimensions vary from several centimeters to micrometers. Organization of about 60,000 conducting airways and 33 million respiratory airways in a limited space results in a complex structure. Due to this inherent complexity and a high number of airways, using target-oriented dimensional reduction is inevitable. In addition, there is no general reduced-order model for various types of problems. This necessitates coming up with an appropriate model from a variety of different reduced-order models to solve the desired... 

Gas exchange, particle deposition, and drug delivery in the human lungs depend on the structure of the human bronchial tree. The conducting airways occupy the main portion of the human lungs and transport air from outside into pulmonary acini where the O2–CO2 exchange with blood occurs. Therefore, the generation of three-dimensional accurate structure of the conducting airways is required for simulation of the transport phenomena in the human respiratory system. The present study proposes an intelligent method for generation of conducting airways based on stochastic parametric Lindenmayer system (L-system). The conducting airways grow into the bronchopulmonary segments simultaneously using... 

The bronchial tree plays a main role in the human respiratory system because the air distribution throughout the lungs and gas exchange with blood occur in the airways whose dimensions vary from several centimeters to micrometers. Organization of about 60,000 conducting airways and 33 million respiratory airways in a limited space results in a complex structure. Due to this inherent complexity and a high number of airways, using target-oriented dimensional reduction is inevitable. In addition, there is no general reduced-order model for various types of problems. This necessitates coming up with an appropriate model from a variety of different reduced-order models to solve the desired...